Himanshu is a ML Engineer with 7+ years of experience in AI/ML development, specializing in deploying models on cloud platforms and integrating solutions into production environments. Vetted by Witarist and ready to join your team within 48 hours.
Direct ownership of AI/ML feature development and deployment from design to production across multiple projects.
Expertise in Python and AI/ML libraries like TensorFlow, Keras, and PyTorch.
Active involvement in code reviews, sanitisation, refactoring, and testing/debugging.
Developed and deployed AI-powered interview functionality for candidate evaluation, integrating backend logic for response evaluation and scoring.
Implemented and optimized Large Language Models (LLMs) like LLAMA, BART, and GPT for tasks such as interview question generation and text summarization.
Engineered and deployed ML platforms for critical applications like predicting suspicious banking activities and malignant/benign tumour identification.
Overview: HireAI is a job portal enabling companies to post job openings and candidates to apply. Responsibilities: Developed microservices using FastAPI to manage job postings, applications, and AI interview scheduling. Designed and implemented chatbot prompting for AI interview processes. Integrated backend logic for AI interviewer functionality, including response evaluation and scoring. Ensured scalability and performance optimization of the application.
Key outcomes:
Developed AI-powered interview functionality for candidate evaluation.
Integrated backend logic for response evaluation and scoring.
Overview: Developed a platform for a banking client to predict suspicious banking activities. Responsibilities: Implemented and trained ML algorithms for suspicious activity prediction. Performed hyper-parameter tuning to optimize model accuracy. Tested and improved the accuracy of the prediction model. Utilized Ray Cluster for parallel processing of ML/non-ML tasks.
Key outcomes:
Developed an ML platform for predicting suspicious banking activities.
Implemented and optimized ML algorithms for accuracy.
ConvAI — interview question answers in programming + LLMs (LLAMA + BART + speech-to-text).
Key outcomes:
Implemented LLAMA, BART, and speech-to-text models for generating interview answers.
Improved model accuracy through hyper-tuning and testing.
Kognetics v2.0.0 — ML-driven business decisions + graphical visualization + E2E Business AI Decision Services.
Key outcomes:
Developed an E2E Business AI Decision Services Management platform.
Deployed trained models on cloud for business decision-making.
Wrote reusable business logic/APIs.
Business Signal Predictor — detecting 500+ business signals (FUNDING + ACQUISITION) + web scraping micro-service for live signals.
Key outcomes:
Developed a web scraping micro-service to collect 500+ business signals.
Trained ML models for signal prediction and stored data in RDF/GraphDB.
Deployed the trained models on cloud.
Himanshu
AI/ML Engineer